Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 7 additions & 3 deletions Lib/tarfile.py
Original file line number Diff line number Diff line change
Expand Up @@ -478,9 +478,13 @@ def __write(self, s):
is ready to be written.
"""
self.buf += s
while len(self.buf) > self.bufsize:
self.fileobj.write(self.buf[:self.bufsize])
self.buf = self.buf[self.bufsize:]
if len(self.buf) > self.bufsize:
# Write whole blocks from buf and keep the rest. Slicing off
# each block would copy the whole remaining buffer every time.
end = (len(self.buf) - 1) // self.bufsize * self.bufsize
for i in range(0, end, self.bufsize):
self.fileobj.write(self.buf[i:i + self.bufsize])
self.buf = self.buf[end:]

def close(self):
"""Close the _Stream object. No operation should be
Expand Down
20 changes: 20 additions & 0 deletions Lib/test/test_tarfile.py
Original file line number Diff line number Diff line change
Expand Up @@ -2977,6 +2977,26 @@ def test_pax_limits(self):

class MiscTest(unittest.TestCase):

def test_stream_write_large_chunks(self):
# A chunk spanning many blocks is still written one block at a time.
data = bytes(range(256)) * 1024
writes = []
class Recorder(io.BytesIO):
def write(self, b):
writes.append(b)
return super().write(b)
fobj = Recorder()
with tarfile.open(fileobj=fobj, mode='w|',
copybufsize=len(data)) as tar:
tarinfo = tarfile.TarInfo('data')
tarinfo.size = len(data)
tar.addfile(tarinfo, io.BytesIO(data))
self.assertEqual({(type(b), len(b)) for b in writes},
{(bytes, tarfile.RECORDSIZE)})
fobj.seek(0)
with tarfile.open(fileobj=fobj, mode='r|') as tar:
self.assertEqual(tar.extractfile(tar.next()).read(), data)

def test_char_fields(self):
self.assertEqual(tarfile.stn("foo", 8, "ascii", "strict"),
b"foo\0\0\0\0\0")
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,2 @@
Speed up writing :mod:`tarfile` streams (``w|`` modes) with a large
*copybufsize*: each chunk is no longer re-copied once per output block.
Loading